This project builds upon our successful recognition of two immunoperoxidase-labeled monoclonal antibodies (Mabs) which have detected tumor-associated antigens expressed on sputum epithelial cells two years in advance of clinical lung cancer (sensitivity 91%, specificity 88%) from participants in the NCI-collaborative Early Lung Cancer detection trial at Johns Hopkins (JHMI) (1). To validate these observations, we have initiated a Collaborative Research and Development Agreement (CRADA) among the NCI, the JHMI, the University of Pennsylvania (U Penn) and Abbott Laboratories (Abbott) specifically to a) refine antibody and immuno- cytochemistry stain procedures; b) coordinate an assessment of screening sensitivity/specificity for second primary lung cancer by Mab detection and binding of tumor-associated antigens, compared with standard Papanicolaou stained sputum cell morphology; and c) recruit 1,000 stage I resected lung cancer patients from 11 university-affiliated oncology programs for specimen production by sputum induction and bronchoalveolar lavage. This application seeks support for three years of patient follow-up, laboratory and statistical procedures to assess the ability of these Mab markers to predict the development of second primary lung cancers in this high-risk population (1-5% expected annual incidence) (2). In essence, the objective of this research is to continue and complete the validation of Mabs to discriminate cells from patients (with completely resected Stage I lung cancer) who will go on to develop a second primary lung cancer from those who will not. Monoclonal antibodies originally made by CRADA tumor biologist (Mulshine) with colleagues at the NCI (3,4), now have been purified and grown in quantity by Abbott for this validation study. the immuno-cytochemical staining of these and other potential markers has been optimized by the CRADA cytopathologist (Gupta) at U Penn (5). Using video- enhanced optical microscopy, the Frost Center laboratory (Tockman) at JHMI has developed quantitative criteria to recognize the presence of these immunostained-antibody labels, which, when combined with morphological features can separate lung cancer cells compared to unlabeled controls (6). JHMI Oncology Biostatistics (Piantadosi) and the JHMI Frost Center (Tockman), respectively, coordinate the flow and analysis of data and specimens between field sites and central laboratories. This study represents the first trial of a biomarker developed to recognize early lung cancer which has made the transition from the tumor biology laboratory to population testing. If validated in the present study, Mabs recognized by quantitative criteria could lead to the development of reproducible lung cancer intermediate endpoints, suitable for marker and intervention efficacy studies in larger populations.